| *** ccrouch has quit IRC | 00:05 | |
| *** ccrouch has joined #openstack-ceilometer | 00:16 | |
| *** _cjones_ has quit IRC | 00:18 | |
| *** _cjones_ has joined #openstack-ceilometer | 00:19 | |
| *** _cjones_ has quit IRC | 00:23 | |
| *** sdake_ has joined #openstack-ceilometer | 00:33 | |
| *** alexpilotti has quit IRC | 00:40 | |
| *** sbfox has joined #openstack-ceilometer | 00:42 | |
| *** shakayumi has quit IRC | 00:45 | |
| *** sbfox has quit IRC | 01:05 | |
| *** fnaval has quit IRC | 01:08 | |
| *** sdake_ has quit IRC | 01:13 | |
| *** rwsu has quit IRC | 01:30 | |
| *** jmanko has quit IRC | 01:32 | |
| *** fnaval has joined #openstack-ceilometer | 01:32 | |
| *** zhiyan|afk has joined #openstack-ceilometer | 01:32 | |
| *** ondergetekende_ has joined #openstack-ceilometer | 01:32 | |
| *** sileht_ has joined #openstack-ceilometer | 01:32 | |
| *** rwsu has joined #openstack-ceilometer | 01:33 | |
| *** theanalyst has quit IRC | 01:33 | |
| *** dougwig has quit IRC | 01:33 | |
| *** ondergetekende has quit IRC | 01:33 | |
| *** zhiyan has quit IRC | 01:33 | |
| *** dougwig has joined #openstack-ceilometer | 01:33 | |
| *** sileht has quit IRC | 01:33 | |
| *** dougwig has quit IRC | 01:33 | |
| *** dougwig has joined #openstack-ceilometer | 01:33 | |
| *** zhiyan|afk is now known as zhiyan | 01:34 | |
| *** theanalyst has joined #openstack-ceilometer | 01:34 | |
| *** openstack has joined #openstack-ceilometer | 01:42 | |
| *** eglynn_ has joined #openstack-ceilometer | 01:42 | |
| *** andreluiz has joined #openstack-ceilometer | 01:42 | |
| *** dougwig has quit IRC | 01:42 | |
| *** theanalyst has quit IRC | 01:42 | |
| *** eglynn has quit IRC | 01:42 | |
| *** dougwig has joined #openstack-ceilometer | 01:42 | |
| *** dekozo has quit IRC | 01:42 | |
| *** dougwig has quit IRC | 01:42 | |
| *** dougwig has joined #openstack-ceilometer | 01:42 | |
| *** larsks|alt has joined #openstack-ceilometer | 01:43 | |
| *** openstackgerrit has joined #openstack-ceilometer | 01:52 | |
| *** nosnos has joined #openstack-ceilometer | 01:52 | |
| *** theanalyst has joined #openstack-ceilometer | 01:52 | |
| *** shakamunyi has joined #openstack-ceilometer | 01:58 | |
| *** sdake_ has joined #openstack-ceilometer | 01:59 | |
| *** sdake has quit IRC | 02:10 | |
| *** larsks|alt is now known as larsks | 02:12 | |
| *** hhuang has joined #openstack-ceilometer | 02:17 | |
| *** harlowja is now known as harlowja_away | 02:21 | |
| *** shakayumi has joined #openstack-ceilometer | 02:24 | |
| *** shakamunyi has quit IRC | 02:27 | |
| *** shardy has quit IRC | 02:32 | |
| *** sdake_ has quit IRC | 02:37 | |
| *** sdake_ has joined #openstack-ceilometer | 02:43 | |
| *** Longgeek has joined #openstack-ceilometer | 02:49 | |
| *** amalagon has quit IRC | 03:06 | |
| *** X019 has quit IRC | 03:37 | |
| *** deepthi has joined #openstack-ceilometer | 03:45 | |
| *** X019 has joined #openstack-ceilometer | 03:50 | |
| *** nosnos has quit IRC | 03:50 | |
| *** nosnos has joined #openstack-ceilometer | 03:51 | |
| *** nosnos has quit IRC | 03:55 | |
| *** renlt has joined #openstack-ceilometer | 04:07 | |
| *** hhuang has quit IRC | 04:24 | |
| *** swamireddy has joined #openstack-ceilometer | 04:25 | |
| *** nosnos has joined #openstack-ceilometer | 04:30 | |
| *** hhuang has joined #openstack-ceilometer | 04:41 | |
| *** yatin has joined #openstack-ceilometer | 04:44 | |
| *** sbfox has joined #openstack-ceilometer | 04:47 | |
| *** ildikov has quit IRC | 04:58 | |
| *** sdake_ has quit IRC | 05:02 | |
| *** amalagon has joined #openstack-ceilometer | 05:07 | |
| *** sileht_ is now known as sileht | 05:38 | |
| *** sileht has joined #openstack-ceilometer | 05:40 | |
| *** amalagon has quit IRC | 05:42 | |
| *** ifarkas has joined #openstack-ceilometer | 05:43 | |
| *** ildikov has joined #openstack-ceilometer | 05:57 | |
| *** cmyster has joined #openstack-ceilometer | 06:00 | |
| *** cmyster has joined #openstack-ceilometer | 06:00 | |
| *** amalagon has joined #openstack-ceilometer | 06:05 | |
| *** zigo has joined #openstack-ceilometer | 06:08 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ceilometer: Imported Translations from Transifex https://review.openstack.org/127723 | 06:10 |
|---|---|---|
| *** Ala has joined #openstack-ceilometer | 06:14 | |
| *** Ala has quit IRC | 06:20 | |
| *** Ala has joined #openstack-ceilometer | 06:20 | |
| *** asalkeld has quit IRC | 06:35 | |
| *** asalkeld has joined #openstack-ceilometer | 06:52 | |
| *** Guest86578 is now known as d0ugal | 07:04 | |
| *** d0ugal has quit IRC | 07:05 | |
| *** d0ugal has joined #openstack-ceilometer | 07:05 | |
| *** ccrouch has quit IRC | 07:08 | |
| *** ccrouch has joined #openstack-ceilometer | 07:09 | |
| *** shardy has joined #openstack-ceilometer | 07:25 | |
| *** eglynn_ has quit IRC | 07:29 | |
| *** eglynn_ has joined #openstack-ceilometer | 07:34 | |
| *** _nadya_ has joined #openstack-ceilometer | 07:35 | |
| *** safchain has joined #openstack-ceilometer | 07:52 | |
| *** liusheng has joined #openstack-ceilometer | 07:58 | |
| *** alexpilotti has joined #openstack-ceilometer | 08:11 | |
| openstackgerrit | Samta Rangare proposed a change to openstack/ceilometer: Alarms listing based on "timestamp" https://review.openstack.org/128573 | 08:14 |
| *** rwsu has quit IRC | 08:31 | |
| *** rwsu has joined #openstack-ceilometer | 08:56 | |
| *** _nadya_ has quit IRC | 09:11 | |
| *** eglynn__ has joined #openstack-ceilometer | 09:13 | |
| *** eglynn_ has quit IRC | 09:14 | |
| *** eglynn__ is now known as eglynn | 09:15 | |
| *** nellysmitt has joined #openstack-ceilometer | 09:23 | |
| *** lsmola has quit IRC | 09:26 | |
| *** admin0 has joined #openstack-ceilometer | 09:32 | |
| *** renlt has quit IRC | 09:34 | |
| *** hhuang has quit IRC | 09:35 | |
| *** lsmola has joined #openstack-ceilometer | 09:42 | |
| *** cdent has joined #openstack-ceilometer | 10:04 | |
| *** Ala has quit IRC | 10:22 | |
| *** Ala has joined #openstack-ceilometer | 10:22 | |
| *** claudiub has joined #openstack-ceilometer | 10:25 | |
| openstackgerrit | Chris Dent proposed a change to openstack/ceilometer: Add a console-script to inspect coordination groups https://review.openstack.org/127271 | 10:42 |
| *** safchain has quit IRC | 10:45 | |
| *** asalkeld has left #openstack-ceilometer | 10:55 | |
| *** sbfox has quit IRC | 11:07 | |
| *** julim has joined #openstack-ceilometer | 11:09 | |
| nellysmitt | eglynn: Hi! Could you please tell what is the current status of this bluprint? https://blueprints.launchpad.net/ceilometer/+spec/neutron-ceilometer-floating-ip-association | 11:34 |
| *** swamireddy1 has joined #openstack-ceilometer | 11:37 | |
| eglynn | jd__: ^^^ that's one of your BPs that nellysmitt is asking about ... do you remember the backgroud to it, or whether the idea is still relevant? | 11:37 |
| nellysmitt | eglynn: it seems that update events are implemented but delete notifications are not handled | 11:38 |
| nellysmitt | https://github.com/openstack/ceilometer/blob/master/ceilometer/network/notifications.py#L56 | 11:39 |
| nellysmitt | should it be fixed? | 11:39 |
| *** swamireddy has quit IRC | 11:40 | |
| eglynn | nellysmitt: you'd need to check if dhellmann's observation in that comment still holds true | 11:42 |
| DinaBelova | eglynn, nellysmitt and I looked on this Bp - it looks like as we're processing ip.floating.update notificaitons, it's fixed already | 11:42 |
| DinaBelova | not sure really :) | 11:42 |
| eglynn | DinaBelova: nellysmitt's question was on the handling of the delete as opposed to update notifications | 11:43 |
| *** _nadya_ has joined #openstack-ceilometer | 11:43 | |
| eglynn | DinaBelova: which aren't handled IIUC, as the notification from neutron was apparently incomplete | 11:43 |
| DinaBelova | eglynn, I guess it's kind of separated quesiotns... | 11:44 |
| eglynn | DinaBelova: a-ha, ok | 11:44 |
| DinaBelova | assosiation in the Bp looks like update | 11:44 |
| DinaBelova | and eleteion is not supported for all neutron notificaitons | 11:44 |
| DinaBelova | deletion* | 11:44 |
| eglynn | nellysmitt: I'd recommend double-checking in the neutron code that the deletion notification is still incomplete | 11:44 |
| *** swamireddy1 has quit IRC | 11:44 | |
| eglynn | nellysmitt: if it is, it would have to be fixed on the neutron side | 11:45 |
| nellysmitt | ok, I'll check | 11:45 |
| eglynn | nellysmitt: (BTW that would be a valid contribution to support your OPW application, patch doesn't necessarily have to apply to the same project repo) | 11:46 |
| DinaBelova | eglynn, that's why we've started this research :) | 11:46 |
| eglynn | DinaBelova: cool | 11:46 |
| nellysmitt | indeed, projects can be different | 11:47 |
| *** safchain has joined #openstack-ceilometer | 11:50 | |
| DinaBelova | nellysmitt, you can try to debug it on your devstack or ask some neutron folks about this moment | 11:50 |
| nellysmitt | good idea, will do :) | 11:50 |
| *** deepthi has quit IRC | 11:51 | |
| eglynn | nellysmitt: OPW application deadline is Oct 22, amiright? | 11:51 |
| nellysmitt | yup | 11:51 |
| nellysmitt | I have a small docco contribution for now | 11:52 |
| nellysmitt | so, in theory I can apply already | 11:52 |
| *** cmyster has quit IRC | 11:52 | |
| *** Ala has quit IRC | 11:52 | |
| nellysmitt | but I want more :D | 11:52 |
| eglynn | nellysmitt: yeah I was gonna suggest getting a "banker" patch landed just as insurance | 11:52 |
| eglynn | nellysmitt: ... but I see you've already gone and done that https://review.openstack.org/#/q/owner:nellysmitt,n,z | 11:53 |
| eglynn | nellysmitt: ... way ahead of me :) | 11:53 |
| nellysmitt | :D | 11:53 |
| _elena_ | ildikov, I want to add some information about cinder-volume-usage-audit to the admin guide, where can I do it? I can't see needed file in ceilo docs, is it in the separated repo or I missed smth? | 11:55 |
| *** mspreitz has joined #openstack-ceilometer | 11:56 | |
| * cdent goes walkies | 12:06 | |
| DinaBelova | nellysmitt, eglynn - fyi - http://paste.openstack.org/show/121232/ and http://paste.openstack.org/show/121233/ shows the difference between notificaiton info for create.start and delete.start notifications - difference is in the payload, actually | 12:07 |
| DinaBelova | for delete there is ID there, for create - name, etc. | 12:07 |
| *** amalagon has quit IRC | 12:08 | |
| *** jdandrea has quit IRC | 12:09 | |
| eglynn | DinaBelova: yeah, isn't that pretty much what dhellmann's comment was saying in the code - different/incomplete metadata in the notification payload? | 12:09 |
| DinaBelova | eglynn, yeah, it looks like so | 12:09 |
| *** jdandrea has joined #openstack-ceilometer | 12:09 | |
| DinaBelova | although I'm not sure | 12:10 |
| DinaBelova | if it's a blocker | 12:10 |
| DinaBelova | and how does it influence of the delete notifications... | 12:10 |
| DinaBelova | I guess I don't understand some metadata moment here | 12:10 |
| eglynn | DinaBelova: I'd imagine dhellmann's reasoning was the last sample for that the neutron resource wouldn't show up in the same query as the previous samples | 12:11 |
| eglynn | DinaBelova: ... as say the project_id is set correctly in the previous samples, but not the last sample derived from the delete notification | 12:12 |
| eglynn | DinaBelova: ... (just using project_id as an example there, subsitute any field you're seeing missing from the delete notification) | 12:12 |
| eglynn | cdent: hola | 12:14 |
| eglynn | cdent: I've got a build of a new openstack-ceilometer-ipmi RPM for fedora | 12:14 |
| eglynn | cdent: ... you wanna give it a whirl on your baremetal setup? | 12:14 |
| eglynn | cdent: (... I'll mail ya the yum repo details) | 12:15 |
| DinaBelova | eglynn, a-ha, cool - I suppose we can suggest nellysmitt to fix the notificaiton format for the *.*.end notificaitons - to add there all the info we're having for the *.*.create + resource id | 12:19 |
| DinaBelova | eglynn - did I get it correclty? | 12:19 |
| *** anna_ has joined #openstack-ceilometer | 12:19 | |
| eglynn | DinaBelova: my understanding is that the info available in the *.delete.end notification payload should match the *.create.end | 12:21 |
| DinaBelova | eglynn, yeah, sorry, wrong phrase | 12:22 |
| DinaBelova | eglynn, you're right | 12:22 |
| *** X019 has quit IRC | 12:22 | |
| eglynn | cool :) | 12:33 |
| *** jmank has joined #openstack-ceilometer | 12:35 | |
| *** anna_ is now known as X019 | 12:41 | |
| cdent | eglynn: I'm back now, yeah, happy to give it a god | 12:42 |
| cdent | go! | 12:42 |
| eglynn | cdent: thank you sir! (... I've mailed repo details offline) | 12:42 |
| *** jmatthews has joined #openstack-ceilometer | 12:42 | |
| cdent | ✔ | 12:43 |
| *** jaypipes has joined #openstack-ceilometer | 12:49 | |
| DinaBelova | eglynn, fyi https://bugs.launchpad.net/neutron/+bug/1381504 | 12:50 |
| eglynn | DinaBelova: thanks! | 12:51 |
| DinaBelova | eglynn, after that nellysmitt will remove that dhellmann's comment and reanable the delete notifications :) | 12:51 |
| DinaBelova | nice :) | 12:51 |
| eglynn | coolness :) | 12:52 |
| *** pradk has joined #openstack-ceilometer | 12:54 | |
| *** claudiub has quit IRC | 12:57 | |
| *** isviridov is now known as isviridov_away | 13:02 | |
| ildikov | _elena_: hi | 13:05 |
| ildikov | _elena_: here you can find the sections of the Telemetry Admin Guide chapter: https://github.com/openstack/openstack-manuals/tree/master/doc/admin-guide-cloud/telemetry | 13:06 |
| DinaBelova | ildikov, _elena_ - it looks like section_telemetry-data-collection.xml seems to be the right place? | 13:06 |
| *** Dafna has quit IRC | 13:06 | |
| _elena_ | ildikov, thanks! | 13:08 |
| _elena_ | DinaBelova, i think yes | 13:08 |
| DinaBelova | щл | 13:08 |
| DinaBelova | ok* | 13:08 |
| *** Ala has joined #openstack-ceilometer | 13:09 | |
| *** gordc has joined #openstack-ceilometer | 13:10 | |
| *** yatin has quit IRC | 13:11 | |
| ildikov | DinaBelova: _elena_: yeap, it looks like that section will be the one | 13:12 |
| openstackgerrit | gordon chung proposed a change to openstack/ceilometer: add script to generate test event data https://review.openstack.org/128445 | 13:13 |
| *** swamireddy has joined #openstack-ceilometer | 13:15 | |
| _elena_ | ildikov, may you help me? I added something to xml, how can i generate this xml to see what happened? | 13:15 |
| ildikov | _elena_: one sec, I will look for the link to the guideline | 13:15 |
| *** IvanBerezovskiy has left #openstack-ceilometer | 13:16 | |
| *** ala_ has joined #openstack-ceilometer | 13:17 | |
| openstackgerrit | gordon chung proposed a change to openstack/ceilometer: refactor hbase storage code https://review.openstack.org/127591 | 13:19 |
| openstackgerrit | gordon chung proposed a change to openstack/ceilometer: refactor hbase storage code https://review.openstack.org/127591 | 13:19 |
| *** Ala has quit IRC | 13:20 | |
| *** alexpilotti has quit IRC | 13:24 | |
| ildikov | _elena_: here it is: https://wiki.openstack.org/wiki/Documentation/HowTo#Download_the_documentation | 13:27 |
| _elena_ | ildikov, thanks! | 13:28 |
| ildikov | _elena_: you will need maven, the guide tells you how to install and what command you need to run | 13:28 |
| ildikov | _elena_: I only had xml formatting issues, nothing specific to maven, so you should be able to make it work easily | 13:29 |
| _elena_ | ildikov, ok, i'll try | 13:32 |
| *** ala_ has quit IRC | 13:32 | |
| sileht | /1/1 | 13:38 |
| *** eglynn has quit IRC | 13:42 | |
| *** eglynn has joined #openstack-ceilometer | 13:43 | |
| *** _nadya_ has quit IRC | 13:49 | |
| *** _nadya_ has joined #openstack-ceilometer | 13:49 | |
| *** _nadya__ has joined #openstack-ceilometer | 13:52 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Add missing alarm options to the documentation https://review.openstack.org/127908 | 13:52 |
| *** _nadya_ has quit IRC | 13:52 | |
| *** sdake_ has joined #openstack-ceilometer | 13:59 | |
| cdent | eglynn this is taking considerably longer than expected: i'm not actually ipmi enabled so I have to hack the ipmitool calls to talk to the remote host | 14:02 |
| cdent | which requires vpn | 14:02 |
| cdent | which tends to irregularly break my network for setting other requirements up | 14:03 |
| * cdent hates vpn | 14:03 | |
| eglynn | cdent: ok, I'll crack on with some other packaging changes in the meantime | 14:03 |
| eglynn | cdent: (RC3 rebase etc.) | 14:04 |
| eglynn | cdent: thanks! | 14:04 |
| cdent | at base level it works as expected: just not yet able to get it to give me samples, but should have that soonish | 14:04 |
| cdent | vpn being on breaks glance, strange | 14:07 |
| *** swamireddy has quit IRC | 14:10 | |
| *** isviridov_away is now known as isviridov | 14:12 | |
| *** shakayumi has quit IRC | 14:12 | |
| *** rbak has joined #openstack-ceilometer | 14:14 | |
| *** sdake has joined #openstack-ceilometer | 14:15 | |
| eglynn | cdent: that is indeed weird (vpn breaking glance) | 14:19 |
| eglynn | cdent: BTW in the latest build I fixed the dumb {neutron|nova|keystone}client errors you would have seen in the agent-ipmi.logs | 14:20 |
| eglynn | (dumb in the sense that we load all the discovery extensions whether they're needed or not) | 14:20 |
| eglynn | (... so we need those clients installed alongside the ipmi agent, even though they'll never be used) | 14:20 |
| cdent | urgh | 14:21 |
| *** alexpilotti has joined #openstack-ceilometer | 14:23 | |
| *** fnaval has quit IRC | 14:27 | |
| *** sbfox has joined #openstack-ceilometer | 14:29 | |
| *** alexpilotti has quit IRC | 14:29 | |
| jd__ | aye | 14:32 |
| *** jergerber has joined #openstack-ceilometer | 14:43 | |
| *** cdent has quit IRC | 14:45 | |
| *** tasdomas` is now known as tasdomas | 14:45 | |
| ildikov | _elena_: just a quick question, if you're still around | 14:45 |
| *** cdent has joined #openstack-ceilometer | 14:45 | |
| ildikov | _elena_: I was wondering if you're docco would fit more to the Cinder config part of the Install guide here: http://docs.openstack.org/trunk/install-guide/install/apt/content/ceilometer-install-cinder.html or not | 14:46 |
| ildikov | _elena_: so I would like to ask your opinion too here, TBH, I wasn't 100% aware of what you plan to add and after reading it, I had this thought in mind | 14:49 |
| *** Longgeek has quit IRC | 14:49 | |
| ildikov | eglynn: hi | 14:49 |
| eglynn | ildikov: hey | 14:50 |
| ildikov | eglynn: as you're my warrior prince regarding to the docco fight, if you have an opnion about the question I've just asked _elena_ about, it's welcomed :) | 14:51 |
| * eglynn reads backlog | 14:52 | |
| ildikov | eglynn: tnx, small question, I'm just not 100% sure this time, so I need opinions :) | 14:52 |
| jdandrea | I'm trying to create a sample and looking at http://docs.openstack.org/developer/ceilometer/webapi/v2.html ... I think I should be using the nomenclature in Sample and not OldSample (e.g., 'meter' and not 'counter_name'). However, even though I use the v2 API samples.create wants counter_name (?). | 14:54 |
| ildikov | eglynn: I've just added you as reviewer to the docco patch I asked about | 14:54 |
| eglynn | ildikov, _elena_: even though that volume audit script is strictly part of the cinder installation/setup, it's really doing work on behalk of ceilometer | 14:54 |
| eglynn | ... so seems more logical to put in in the telemetry data collection section | 14:54 |
| ildikov | eglynn: _elena_: ok, I had the other viewpoit in mind, so that what has to be done to get data out of Cinder | 14:56 |
| eglynn | jdandrea: are you POSTing to /v2/meters/meter_name ? | 14:57 |
| eglynn | jdandrea: ... the docco is pretty clear that this expects the OldSample format | 14:58 |
| jdandrea | eglynn: I'm using the python bindings. | 14:58 |
| ildikov | eglynn: _elena_: ... so that what should be configured/used | 14:58 |
| jdandrea | ceclient = ceilometerclient.client.get_client(2, **cecreds) | 14:58 |
| jdandrea | eglynn: Ohhhh. I saw "deprecated in favor of" and immediately thought I should start using Sample. | 14:59 |
| ildikov | eglynn: _elena_: but well the other viewpoint sounds logical too :) | 14:59 |
| DinaBelova | ildikov, eglynn - _elena_ is afk :) | 15:00 |
| eglynn | ildikov: TBH I don't have a strong opinion, that was purely a gut reaction | 15:00 |
| DinaBelova | she'll go through your messages asap | 15:00 |
| ildikov | DinaBelova: I thought so, but anyway, she can read the logs later, it's not that pretty urgent anyway | 15:01 |
| DinaBelova | :) | 15:01 |
| ildikov | DinaBelova: I just had a question in mind and as the patch is in the docco repo, I thought it's better to quickly ask here | 15:01 |
| ildikov | DinaBelova: so if she answers later or tomorrow, that should be fine also | 15:03 |
| DinaBelova | ildikov, a-ha, ok | 15:03 |
| ildikov | eglynn: I don't have a strong opinion either, that is why I thought to ask | 15:03 |
| ildikov | eglynn: as an experience, I'm lazy to read docco, so it seemed a bit more logical to me to point the user to what options need to be used and set in order to be able to collect data in one place, but it's still more a feeling, than a strong opinion | 15:08 |
| ildikov | eglynn: anyway, thanks for sharing your thoughts, I will think this over once more and then put my decision on gerrit :) | 15:09 |
| eglynn | ildikov: sure, how about putting the detail in the cinder section and a cross-section link to that from the telemetry section? | 15:09 |
| ildikov | eglynn: I always forget about the cross section link option, but that sounds a good idea to me actually | 15:10 |
| eglynn | cool | 15:10 |
| ildikov | eglynn: thanks, I will add a comment on gerrit then | 15:12 |
| cdent | eglynn: i'm sad to report that I have found a bug in the ipmi agent code | 15:13 |
| eglynn | cdent: rootwrap sudoers? | 15:13 |
| cdent | nope | 15:13 |
| cdent | if it can't parse a single sensor, it drops all the samples | 15:13 |
| eglynn | cdent: one fails, they all fail? | 15:14 |
| cdent | yes | 15:14 |
| cdent | i'm just confirming the fix now (to see if my analysis is correct) | 15:14 |
| eglynn | cdent: interesting, which one is failing? | 15:14 |
| cdent | oh, all of them | 15:14 |
| cdent | because the code is supposed to ignore stuff it doesn't understand | 15:15 |
| cdent | and then carry on with stuff it does | 15:15 |
| cdent | but there will always be at least one thing that doesn't make sense | 15:15 |
| *** alexpilotti has joined #openstack-ceilometer | 15:15 | |
| eglynn | cdent: a-ha, k ... can you file a bug once you've confirmed your RCA? | 15:15 |
| * cdent nods | 15:15 | |
| eglynn | cdent: ... at this stage we'll prolly have to just carry the fix on stable/juno :( | 15:16 |
| cdent | I reckon even after this is fixed it would be very wise to do some testing in a correct hardware an installation sense | 15:16 |
| cdent | because I've got a lot of bailing wire here | 15:16 |
| eglynn | cdent: BTW I noticed that I'd hosed the rootwrap config in that package, rebuilding currently | 15:16 |
| cdent | yeah, I'm working around that at the moment | 15:17 |
| cdent | that's one of the things that I think a "proper install" will confirm | 15:17 |
| *** fnaval has joined #openstack-ceilometer | 15:17 | |
| cdent | \o/ I have samples! | 15:17 |
| cdent | so the bug is real and the fix is right | 15:17 |
| cdent | meh | 15:17 |
| eglynn | cdent: cool, let's get the fix landed on master at least | 15:18 |
| cdent | this particular set of code has not been our finest hour :( | 15:18 |
| eglynn | yeap, agreed | 15:19 |
| *** sbfox has quit IRC | 15:20 | |
| eglynn | cdent: BTW that package with fixed rootwrap sudoers is now up if you wanna pick latest | 15:25 |
| eglynn | (yum remove -y openstack-ceilometer-ipmi ; yum --enablerepo=copr-juno-testing clean metadata ; yum install -y openstack-ceilometer-ipmi) | 15:26 |
| eglynn | cdent: despite the bug you've found, I'm inclined to push the packaging updates at least | 15:26 |
| eglynn | (since the sources are still based on RC2 in any case) | 15:26 |
| *** nosnos has quit IRC | 15:27 | |
| cdent | makes sense, i'll try that in a mo | 15:27 |
| cdent | there's the bug eglynn : https://bugs.launchpad.net/ceilometer/+bug/1381600 | 15:29 |
| cdent | will push the fix pronto | 15:29 |
| eglynn | cdent: thank you sir! | 15:29 |
| cdent | ugh | 15:29 |
| cdent | I think I've just found another bug | 15:29 |
| cdent | (of the same sort) | 15:30 |
| eglynn | darn! | 15:30 |
| cdent | (this is _exactly_ why I used full data sets in the tests for the similar ironic stuff, rather than the smaller samples used in this code) | 15:30 |
| * cdent will remember to stick more strongly to guns in the future | 15:30 | |
| cdent | the second issue is here: | 15:31 |
| eglynn | cdent: yep ... and feel free to modify the tests in your patch to the style you'd in mind in the initial reviews | 15:31 |
| cdent | https://github.com/openstack/ceilometer/blob/0820c95f0d335e978a92f5b65848d51623e4df78/ceilometer/ipmi/pollsters/sensor.py#L56 | 15:32 |
| cdent | that exception will blow out the entire collection of samples from the current ipmi call | 15:32 |
| cdent | instead of just skipping one that is bad | 15:32 |
| cdent | (the original issue is at line 65 | 15:32 |
| eglynn | crapski :( | 15:34 |
| eglynn | well at least we're finding these issue now before releasing packages | 15:34 |
| cdent | I think the right fix is log instead of raise | 15:34 |
| eglynn | (though of course would have been better to have gotten the fixes into RC3) | 15:34 |
| eglynn | agree | 15:35 |
| *** isviridov is now known as isviridov_away | 15:39 | |
| *** packet has joined #openstack-ceilometer | 15:46 | |
| cdent | jd__: are you expecting/hoping everyone to migrate to gnocchi as soon as it is ready? (re your comment on https://review.openstack.org/#/c/128391/ ) | 15:47 |
| jd__ | cdent: well the point is to finish and integrate completely during Kilo and obsolete the current sample scheme | 15:50 |
| jd__ | cdent: so I don't see putting efforts into that BP during Kilo being a smart idea | 15:50 |
| jd__ | though the idea itself sounds good to me | 15:51 |
| cdent | but presumably the old way will have to have the usual long deprecation cycle | 15:51 |
| jd__ | agreed | 15:51 |
| cdent | and people's existing installations might like some expiration granularity as they struggle into the future | 15:52 |
| jd__ | but is it worthwile adding features in K for something deprecating in L? | 15:52 |
| * cdent doesn't know, thus the query | 15:52 | |
| jd__ | well | 15:52 |
| jd__ | best thing is that it's going to be deprecated in K | 15:52 |
| jd__ | so "here's K with new features for samples that are now being deprecated" | 15:52 |
| jd__ | O_O | 15:53 |
| cdent | Assuming things go to plan, I guess, yeah. | 15:54 |
| *** ildikov has quit IRC | 15:56 | |
| *** _cjones_ has joined #openstack-ceilometer | 16:03 | |
| eglynn | cdent: re. severity of https://bugs.launchpad.net/ceilometer/+bug/1381600 ... High d'ya agree? | 16:05 |
| cdent | yes | 16:05 |
| openstackgerrit | Chris Dent proposed a change to openstack/ceilometer: Handle poorly formed individual sensor readings https://review.openstack.org/128694 | 16:23 |
| cdent | there go eglynn | 16:23 |
| eglynn | cdent: thank you sir! | 16:23 |
| *** admin0 has quit IRC | 16:24 | |
| eglynn | llu, jd__, gordc, sileht, DinaBelova, _nadya__: I'd appreciate quick eyes on https://review.openstack.org/128694 if you got a minute | 16:28 |
| eglynn | thanks! :) | 16:28 |
| cdent | eglynn: the rootwrap changes appear to be okay | 16:33 |
| eglynn | cdent: nice, thanks for confirming | 16:33 |
| gordc | cdent: how do you feel about inheriting TestTemperatureSensorPollster rather than TestPollsterBase | 16:37 |
| *** yatin has joined #openstack-ceilometer | 16:38 | |
| cdent | gordc: I just went with replicating the pattern that was already there | 16:38 |
| gordc | cdent: good enough for me. +A | 16:38 |
| cdent | (which was fairly verbose test per thingie) | 16:38 |
| gordc | i'm very firm with my opinions. :) | 16:38 |
| *** sbfox has joined #openstack-ceilometer | 16:39 | |
| cdent | In another context I'd be full of despair and hate for those tests. | 16:39 |
| *** safchain has quit IRC | 16:39 | |
| cdent | But I figured it was better just to get it done | 16:39 |
| _nadya__ | cdent: only one question: shouldn't we add logging in case of error or it's not useful for chunk lost? | 16:39 |
| gordc | cdent: fair enoug | 16:40 |
| openstackgerrit | gordon chung proposed a change to openstack/ceilometer: add script to generate test event data https://review.openstack.org/128445 | 16:40 |
| cdent | As I said in the commit message _nadya_ there are a _lot_ skipped chunks, logging them would be extremely verbose | 16:40 |
| cdent | and it is an expected condition | 16:40 |
| cdent | what wasn't expected was that the condition should make everything die | 16:40 |
| _nadya__ | cdent: I see, sorry I missed that. Ok! Agreed with approval :) | 16:41 |
| cdent | :) | 16:41 |
| *** safchain has joined #openstack-ceilometer | 16:41 | |
| *** sbfox has quit IRC | 16:43 | |
| *** sbfox has joined #openstack-ceilometer | 16:43 | |
| mspreitz | I have a stupid question about the CLI command `ceilometer resource-list -q <QUERY>`. What does it mean? I mean, how is the QUERY relevant? | 16:44 |
| eglynn | mspreitz: it allows you to constrain the query, e.g. ceilometer resource-list -q "project=PROJECT_ID" | 16:45 |
| mspreitz | but what exactly does the constraint mean? Is it a filter on Samples? If so, how does that affect the answer to my request for resources? | 16:46 |
| mspreitz | eglynn: I suspect the answer is this: `ceilometer resource-list -q QUERY` returns all the resources mentioned in Samples that mach the QUERY. Have I got it right? | 16:48 |
| mspreitz | s/mach/match/ | 16:48 |
| eglynn | mspreitz: interesting question :) ... it actually depends on the storage backend | 16:49 |
| mspreitz | eglynn: ruh roh! | 16:50 |
| eglynn | mspreitz: lemme fill out some background to explain it better | 16:50 |
| mspreitz | eglynn: great, thanks! | 16:51 |
| eglynn | mspreitz: so originally all resources queries where essentially queries over the samples in the mongodb case | 16:51 |
| eglynn | mspreitz: this was because the resource representation includes a first and last sample timestamp | 16:52 |
| eglynn | mspreitz: but the performance absolutely sucked, as the entire set of samples had to be walked to satisfy the query | 16:52 |
| eglynn | mspreitz: so I rationalized that approach with https://github.com/openstack/ceilometer/commit/7b4d4d44650bbe10225084b3ab5155f70ebebb95 | 16:53 |
| *** amalagon has joined #openstack-ceilometer | 16:53 | |
| eglynn | mspreitz: the commit message describes the compromise approach taken | 16:53 |
| eglynn | mspreitz: i.e. general resource queries ==> satisfied from the much smaller resources collection | 16:54 |
| eglynn | mspreitz: i.e. time-constrained resource queries ==> satisfied from the meter collection (i.e. the raw samples) | 16:54 |
| mspreitz | yep, just read it | 16:54 |
| *** amalagon_ has joined #openstack-ceilometer | 16:55 | |
| mspreitz | but it sounds like that change only affects implementation, not the behavior delivered to clients | 16:55 |
| mspreitz | Does that change affect the results delivered to clients? | 16:56 |
| mspreitz | eglynn: that looks like an implementation change not an API change, not a behavior change. Is that right? | 16:57 |
| eglynn | the first and last timestamps samples may not be set accurately in the results for older resources | 16:57 |
| mspreitz | eglynn: OK. But in terms of the results to queries for resources, this change is a no-op, right? | 16:58 |
| eglynn | mspreitz: by older resources, I mean resources that existed in the metering store prior to that patch | 16:58 |
| *** amalagon has quit IRC | 16:58 | |
| eglynn | mspreitz: the same resources are returned, if that's what you mean | 16:58 |
| mspreitz | eglynn: or maybe those timestamps are in the query results, I guess that is why you bring it up | 16:58 |
| mspreitz | eglynn: yes, my concern here is for which resources are returned | 16:59 |
| eglynn | mspreitz: yes, the first & last timestamps are included representation of those resources, that's why I mentioned it | 16:59 |
| *** amalagon_ has quit IRC | 16:59 | |
| eglynn | mspreitz: but the set of resources returned would be essentially the same | 16:59 |
| mspreitz | eglynn: so, if I understand correctly, `ceilometer resource-list -q QUERY` returns a set of resources that does not depend on the backend, although the timestamp bounds in those resources may depend on the backend | 17:00 |
| eglynn | mspreitz: yes | 17:00 |
| mspreitz | eglynn: Let me ask further which resources are returned. Is it exactly those for which there is at least one Sample that matches the query? | 17:00 |
| eglynn | mspreitz: conceptually yes ... but as I said earlier, the query is satisified from a different collection in reality (in most cases) | 17:02 |
| mspreitz | eglynn: but that different collection is maintained automatically and consistently, right? | 17:02 |
| eglynn | mspreitz: yes, it's updated as each sample is inserted | 17:02 |
| mspreitz | OK, thanks. | 17:03 |
| mspreitz | eglynn: next question: is this semantics described in any document? | 17:03 |
| *** _nadya__ has quit IRC | 17:03 | |
| eglynn | mspreitz: to that level of detail? no, I'd suspect not | 17:04 |
| *** amalagon has joined #openstack-ceilometer | 17:04 | |
| *** julim has quit IRC | 17:04 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Avoid clobbering existing class definition https://review.openstack.org/128367 | 17:04 |
| mspreitz | eglynn: Ugh. Next question. I am having trouble writing a resource-list query that filters on meter name. Can you give me an example that works? | 17:04 |
| openstackgerrit | A change was merged to openstack/ceilometer: Clarify deprecation comment to be accurate https://review.openstack.org/128353 | 17:04 |
| mspreitz | eglynn: `ceilometer --debug sample-list -m my-favorite-meter` reveals that in the API, the term is "counter_name" rather than "meter name". But queries on that give me 400 | 17:06 |
| mspreitz | as do queries on counter, meter, and meter_name | 17:06 |
| eglynn | mspreitz: one sec | 17:06 |
| *** yatin has quit IRC | 17:08 | |
| *** harlowja_away is now known as harlowja | 17:15 | |
| mspreitz | eglynn: are you still there? | 17:16 |
| *** _nadya_ has joined #openstack-ceilometer | 17:16 | |
| eglynn | mspreitz: yes | 17:16 |
| *** _cjones_ has quit IRC | 17:17 | |
| *** _cjones_ has joined #openstack-ceilometer | 17:17 | |
| eglynn | mspreitz: sorry I had to finish something else | 17:19 |
| *** _nadya_ has quit IRC | 17:19 | |
| eglynn | mspreitz: k, so there are certain fields that be used to constrain queries on certain endpoints | 17:19 |
| eglynn | mspreitz: for resources the list includes 'project', 'user', 'resource' & 'source' | 17:20 |
| mspreitz | eglynn: is the list for resources documented anywhere? | 17:20 |
| eglynn | mspreitz: it appears in the error message IIRC | 17:20 |
| mspreitz | Yes! DEBUG (client) RESP BODY: {"error_message": {"debuginfo": null, "faultcode": "Client", "faultstring": "Unknown argument: \"counter_name\": unrecognized field in query: [<Query u'counter_name' eq u'network.services.lb.member' >], valid keys: ['end_timestamp', 'end_timestamp_op', 'metaquery', 'pagination', 'project', 'resource', 'source', 'start_timestamp', 'start_timestamp_op', 'user']"}} | 17:21 |
| mspreitz | what's that metaquery thing? | 17:22 |
| eglynn | query on the metadata | 17:23 |
| mspreitz | eglynn: here is an example Sample I got from `ceilometer --degut sample-list`: {"counter_name": "network.services.lb.member", "user_id": null, "resource_id": "5c2caff6-1ad9-4806-9455-98ab524d6fc7", "timestamp": "2014-10-15T03:55:37.842352", "recorded_at": "2014-10-15T03:55:37.858280", "message_id": "1f93506c-541f-11e4-b0d0-fa6363248f18", "source": "openstack", "counter_unit": "member", "counter_volume": 0.0, "project_id" | 17:24 |
| mspreitz | dang, truncated by freenode | 17:25 |
| mspreitz | "project_id": "ba0ddcea2e0744a3aafe3ff26fd6d292", "resource_metadata": {"status": "INACTIVE", "status_description": "None", "weight": "1", "admin_state_up": "True", "pool_id": "c7931134-37f5-430b-aabb-b3953f3ad83a", "address": "10.0.0.212", "protocol_port": "80"}, "counter_type": "gauge"} | 17:25 |
| mspreitz | that's the rest | 17:25 |
| mspreitz | eglynn: so metaquery would be a query over the "resource_metadata" part? | 17:26 |
| eglynn | yep | 17:27 |
| *** changbl has joined #openstack-ceilometer | 17:30 | |
| *** r-daneel has joined #openstack-ceilometer | 17:37 | |
| *** Ala has joined #openstack-ceilometer | 17:40 | |
| *** julim has joined #openstack-ceilometer | 17:42 | |
| *** changbl has quit IRC | 17:43 | |
| *** Ala has quit IRC | 17:48 | |
| *** _nadya_ has joined #openstack-ceilometer | 17:49 | |
| mspreitz | eglynn: this is very surprising. You are telling me that in `ceilometer resource-list -q QUERY`, the query can NOT discriminate on counter_name (AKA meter name) and the query can NOT discriminate based on counter_volume. Have I got that right? | 17:49 |
| eglynn | mspreitz: yes, this is a query over the resources as I explained above, as opposed to the samples, so it cannot be constrained by the sample volume | 17:54 |
| eglynn | mspreitz: there might be dozens of different meters associated with a resource and potentially million of different sample volumes | 17:55 |
| mspreitz | eglynn: OK, so the QUERY in `ceilometer resource-list -q QUERY` is semantically constrained to be about only the Resources, not their associated Samples | 17:55 |
| eglynn | mspreitz: yes, modulo the issue that I mentioned right at the very start (around first & last samples timestamps) | 17:56 |
| mspreitz | eglynn: `ceilometer sample-list -m METER -q QUERY`, OTOH, does get to look at Samples | 17:56 |
| mspreitz | eglynn: right, understand about the time bound subtlety | 17:57 |
| eglynn | mspreitz: yes, sample-list looks at the samples, that's its purpose | 17:57 |
| * cdent dines | 17:58 | |
| mspreitz | eglynn: however, `ceilometer sample-list -m METER -q QUERY` can not do something like GROUP BY, so I can not use it to find resources that have Sample meeting some criterion. I have to fetch all the matching Samples and then reduce the data myself | 17:58 |
| *** _nadya_ has quit IRC | 17:58 | |
| mspreitz | eglynn: have I got that right? | 18:01 |
| *** admin0 has joined #openstack-ceilometer | 18:01 | |
| eglynn | mspreitz: one sec | 18:02 |
| eglynn | mspreitz: sorry, yeah, groupby is intended for the statistics API as there's aggregation going on | 18:09 |
| eglynn | mspreitz: whereas sample-lists reports the raw samples (i.e. no aggregation) | 18:09 |
| eglynn | mspreitz: see http://docs.openstack.org/developer/ceilometer/webapi/v2.html | 18:09 |
| eglynn | mspreitz: ... I'm gonna have to drop off v. soon, after 7pm here | 18:10 |
| amalagon | jd__: looks like the response_dict thing was indeed a bug: https://bugs.launchpad.net/swift/+bug/1381304 | 18:10 |
| jd__ | amalagon: well done! | 18:11 |
| amalagon | :) | 18:11 |
| mspreitz | eglynn: OK, let me try two quick questions | 18:12 |
| mspreitz | eglynn: First, for the CLI, how do I put a datatype in a query? I mean, if I have to compare with something that is not a string? | 18:13 |
| mspreitz | eglynn: Second, does Ceilometer have any way to query for the stats aggregated over all of a period that the client passes in as a paremeter to the API operation I am looking for? | 18:13 |
| mspreitz | Okay, anybody can answer. | 18:15 |
| mspreitz | What do I write on the command line to, e.g., compare counter_volume with a number? | 18:15 |
| *** jergerber has quit IRC | 18:18 | |
| eglynn | mspreitz: I'll be back in a hour, I'll answer your questions then | 18:20 |
| eglynn | (dinner time ... eating) | 18:20 |
| *** eglynn is now known as eglynn-afk | 18:20 | |
| mspreitz | eglynn: thanks | 18:20 |
| *** admin0 has quit IRC | 18:20 | |
| *** _nadya_ has joined #openstack-ceilometer | 18:30 | |
| *** _nadya_ has quit IRC | 18:39 | |
| *** mspreitz has quit IRC | 18:42 | |
| *** jmanko has joined #openstack-ceilometer | 18:52 | |
| *** _nadya_ has joined #openstack-ceilometer | 18:52 | |
| *** mikehn has quit IRC | 18:52 | |
| *** mikehn_ has joined #openstack-ceilometer | 18:52 | |
| *** jmank has quit IRC | 18:53 | |
| *** packet has quit IRC | 18:55 | |
| *** amalagon has quit IRC | 18:55 | |
| *** packet has joined #openstack-ceilometer | 18:57 | |
| *** jaypipes has quit IRC | 18:57 | |
| *** mspreitz has joined #openstack-ceilometer | 19:01 | |
| *** _cjones_ has quit IRC | 19:02 | |
| *** _cjones_ has joined #openstack-ceilometer | 19:03 | |
| *** drjones has joined #openstack-ceilometer | 19:06 | |
| *** _cjones_ has quit IRC | 19:06 | |
| *** tongli has joined #openstack-ceilometer | 19:09 | |
| *** jaypipes has joined #openstack-ceilometer | 19:11 | |
| *** _nadya_ has quit IRC | 19:14 | |
| *** admin0 has joined #openstack-ceilometer | 19:21 | |
| eglynn-afk | mspreitz: I'm back (briefly) | 19:21 |
| *** eglynn-afk is now known as eglynn | 19:21 | |
| *** admin0 has quit IRC | 19:21 | |
| eglynn | mspreitz: "how do I put a datatype in a query" ==> just prefix the value with type:: | 19:22 |
| eglynn | where type is integer, float, string, boolean etc. | 19:22 |
| eglynn | e.g. ceilometer sample-list -m instance -q 'metadata.status==string::shutoff' | 19:22 |
| mspreitz | so, e.g., counter_volume<integer:0.1 | 19:22 |
| mspreitz | eglynn: right? | 19:22 |
| eglynn | no | 19:22 |
| mspreitz | oh no | 19:22 |
| mspreitz | its double colon instead of single | 19:23 |
| eglynn | yep | 19:23 |
| mspreitz | thanks | 19:23 |
| mspreitz | got it | 19:23 |
| eglynn | mspreitz: "does Ceilometer have any way to query for the stats aggregated over all of a period" ==> yep the -q can be timestamp-bounded to constrain to a period | 19:23 |
| eglynn | e.g. ceilometer statistics -m cpu_util -q 'timestamp>START;timestamp<=END' | 19:23 |
| eglynn | where START, END are in ISO8601 format | 19:24 |
| mspreitz | eglynn: what if I want one answer, covering the whole period, not many answers, each for a distinct subperiod? | 19:24 |
| eglynn | mspreitz: the -p option controls the sub-period lenght | 19:24 |
| mspreitz | eglynn: alignment is my question | 19:24 |
| eglynn | mspreitz: defaults to the entire duration if missing | 19:25 |
| *** admin0 has joined #openstack-ceilometer | 19:25 | |
| *** admin0 has quit IRC | 19:25 | |
| mspreitz | eglynn: in the statistics operation, with -p N, are boundaries implicitly set to multiples of N from 1970.0, or multiple of N from time of query op, or ... ? | 19:26 |
| *** cdent has quit IRC | 19:28 | |
| eglynn | mspreitz: well if I understand your question ... | 19:29 |
| eglynn | mspreitz: ... the duration of the selected samples (as constrained by the query condition) are split from the start of that duration by period | 19:29 |
| eglynn | mspreitz: ... NOT from 1970 | 19:29 |
| eglynn | mspreitz: ... in particular we don't report on periods for which we've no datapoints | 19:30 |
| mspreitz | eglynn: I confuse period and duration. Suppose I say `ceilometer statistics -m METER -p N`. That gives me records, each one covering a range of time [start, end); I am speaking of the queried time, not the data found. Successive starts are separated by a multiple of N. Each start is a multiple of N added to or subtracted from some baseline. What is the baseline? | 19:32 |
| eglynn | mspreitz: ok, let me tie down terminology to lift the confusion | 19:33 |
| eglynn | mspreitz: lets say "span" is the entire range of the timestamps covered by the query | 19:33 |
| eglynn | mspreitz: lets say "period" is the individual time-buckets into which these datapoints are aggregated | 19:34 |
| *** safchain has quit IRC | 19:34 | |
| mspreitz | a simple experiment reminded me of duration vs period | 19:34 |
| mspreitz | and I got span | 19:35 |
| eglynn | mspreitz: and finally lets save "duration" in our new, clearer terminology to mean the range of timestamps of actual datapoints that fall *with* each period | 19:35 |
| mspreitz | right | 19:35 |
| mspreitz | so my question is about the periods | 19:35 |
| mspreitz | each starts at some multiple of N added to or subtracted from <what>? | 19:35 |
| eglynn | mspreitz: so if I submit a query like this -q 'timestamp>START;timestamp<=END' | 19:35 |
| eglynn | mspreitz: that selects all match datapoints between START and END | 19:36 |
| mspreitz | that is the span | 19:36 |
| eglynn | mspreitz: exactly, that's the span | 19:36 |
| eglynn | mspreitz: the span is then spilt into periods | 19:36 |
| eglynn | mspreitz: the split occurs from the start of the span | 19:36 |
| mspreitz | great, thanks | 19:36 |
| mspreitz | BTW, how are the periods set when the query does not include START? | 19:37 |
| eglynn | mspreitz: first matching sample timestamp | 19:37 |
| mspreitz | OK, thanks | 19:37 |
| mspreitz | Is this documented anywhere? | 19:38 |
| eglynn | mspreitz: not as completely as it should be, though patch as always are welcome :) | 19:38 |
| eglynn | mspreitz: any follow-up questions, catch me on IRC tomorrow | 19:38 |
| eglynn | mspreitz: (my TZ is GMT, getting late now, so dropping off shortly) | 19:39 |
| mspreitz | ceilometer sample-list -m cpu_util -q "counter_volume<float::0.1" | 19:39 |
| mspreitz | eglynn: gave me 400 | 19:39 |
| mspreitz | HTTP error code 400 | 19:39 |
| mspreitz | eglynn: thanks for all the help so far | 19:40 |
| mspreitz | eglynn: ceilometer sample-list -m cpu_util -q "counter_volume<integer::1" also fails w HTTP return code 400 | 19:42 |
| mspreitz | also no good for real or int | 19:42 |
| openstackgerrit | A change was merged to openstack/ceilometer: refactor hbase storage code https://review.openstack.org/127591 | 19:43 |
| eglynn | mspreitz: as I thought I explained earlier, each query has a list fields that it can be constrained on | 19:43 |
| eglynn | mspreitz: list *of fields | 19:44 |
| eglynn | mspreitz: user, project, resource etc. and then free-form metadata query | 19:44 |
| eglynn | mspreitz: volume is not in that list | 19:44 |
| mspreitz | eglynn: oh, right, sorry, I forgot to ask for the list for sample-list | 19:44 |
| eglynn | mspreitz: just look at the error message, I believe it lists them | 19:45 |
| mspreitz | eglynn: yep, it's there | 19:45 |
| mspreitz | eglynn: so I can not use `ceilometer sample-list` to find samples with extreme values. And I can not use `ceilometer resource-list` to find resource with samples from a given meter. | 19:47 |
| mspreitz | eglynn: there there any plan to enable such things? | 19:47 |
| openstackgerrit | A change was merged to openstack/ceilometer: Handle poorly formed individual sensor readings https://review.openstack.org/128694 | 19:47 |
| mspreitz | s/there there/are there/ | 19:47 |
| *** cdent has joined #openstack-ceilometer | 19:48 | |
| *** eglynn has quit IRC | 19:49 | |
| *** eglynn has joined #openstack-ceilometer | 19:51 | |
| *** cdent has quit IRC | 19:53 | |
| eglynn | mspreitz: ack & ack to your first two assertions | 19:53 |
| eglynn | mspreitz: no specific plans ... but if you have a requirement for these, the usual approach would be file a bug | 19:54 |
| eglynn | mspreitz: and as always, patches are welcome :) | 19:54 |
| mspreitz | eglynn: OK, thanks. Maybe just one final question. If I were to attempt to add the doc that I asked about but is not there, where should it go? | 19:54 |
| eglynn | mspreitz: here's where the API doc should be expanded https://github.com/openstack/ceilometer/blob/master/doc/source/webapi/v2.rst | 19:56 |
| mspreitz | eglynn: thank you VERY much | 19:57 |
| eglynn | mspreitz: np! | 19:57 |
| *** eglynn has quit IRC | 19:57 | |
| *** ifarkas has quit IRC | 19:57 | |
| *** amalagon has joined #openstack-ceilometer | 19:57 | |
| *** tongli has quit IRC | 20:10 | |
| *** tongli has joined #openstack-ceilometer | 20:11 | |
| *** nellysmitt has quit IRC | 20:37 | |
| *** nellysmitt has joined #openstack-ceilometer | 20:38 | |
| *** nellysmitt has quit IRC | 20:42 | |
| *** promulo_ has joined #openstack-ceilometer | 20:55 | |
| *** flwang has joined #openstack-ceilometer | 21:22 | |
| *** mspreitz has quit IRC | 21:23 | |
| *** r-daneel has quit IRC | 21:28 | |
| *** tongli has quit IRC | 21:32 | |
| *** gordc has quit IRC | 21:41 | |
| *** drjones has quit IRC | 21:48 | |
| *** _cjones_ has joined #openstack-ceilometer | 21:49 | |
| *** safchain has joined #openstack-ceilometer | 21:52 | |
| *** _cjones_ has quit IRC | 21:53 | |
| *** packet has quit IRC | 21:58 | |
| *** _cjones_ has joined #openstack-ceilometer | 22:01 | |
| *** asalkeld has joined #openstack-ceilometer | 22:12 | |
| *** sbfox1 has joined #openstack-ceilometer | 22:14 | |
| flwang | greetings, is there any known issue about running tox against icehouse branch? | 22:16 |
| *** sbfox has quit IRC | 22:16 | |
| flwang | I always get this error "Non-zero exit code (2) from test listing." | 22:16 |
| *** sbfox has joined #openstack-ceilometer | 22:17 | |
| *** sbfox1 has quit IRC | 22:19 | |
| *** jmatthews has quit IRC | 22:25 | |
| *** alexpilotti has quit IRC | 22:32 | |
| *** ryanpetrello has quit IRC | 22:34 | |
| *** ryanpetrello has joined #openstack-ceilometer | 22:35 | |
| openstackgerrit | Srinivas Sakhamuri proposed a change to openstack/ceilometer: Allow collector service database connection retry https://review.openstack.org/127128 | 22:50 |
| *** asalkeld has quit IRC | 22:56 | |
| *** promulo_ has quit IRC | 23:01 | |
| *** rbak has quit IRC | 23:04 | |
| *** BernardVDW has joined #openstack-ceilometer | 23:10 | |
| *** asalkeld has joined #openstack-ceilometer | 23:13 | |
| *** _cjones_ has quit IRC | 23:24 | |
| *** _cjones_ has joined #openstack-ceilometer | 23:24 | |
| *** drjones has joined #openstack-ceilometer | 23:26 | |
| *** _cjones_ has quit IRC | 23:29 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ceilometer: Updated from global requirements https://review.openstack.org/127758 | 23:40 |
| *** jaypipes has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!